TheNeuralBit opened a new issue, #23953:
URL: https://github.com/apache/beam/issues/23953
### What would you like to happen?
Currently, we yield an error like this:
```
Traceback (most recent call last):
File "simple_numeric.py", line 71, in <module>
batched_circle_montecarlo(100000)
File "simple_numeric.py", line 60, in batched_circle_montecarlo
p
File
"/tmp/tmp.PH23v4EA7w/venv/lib/python3.8/site-packages/apache_beam/pvalue.py",
line 137, in __or__
return self.pipeline.apply(ptransform, self)
File
"/tmp/tmp.PH23v4EA7w/venv/lib/python3.8/site-packages/apache_beam/pipeline.py",
line 709, in apply
pvalueish_result = self.runner.apply(transform, pvalueish, self._options)
File
"/tmp/tmp.PH23v4EA7w/venv/lib/python3.8/site-packages/apache_beam/runners/runner.py",
line 185, in apply
return m(transform, input, options)
File
"/tmp/tmp.PH23v4EA7w/venv/lib/python3.8/site-packages/apache_beam/runners/runner.py",
line 215, in apply_PTransform
return transform.expand(input)
File
"/tmp/tmp.PH23v4EA7w/venv/lib/python3.8/site-packages/apache_beam/transforms/core.py",
line 1580, in expand
self.infer_batch_converters(pcoll.element_type)
File
"/tmp/tmp.PH23v4EA7w/venv/lib/python3.8/site-packages/apache_beam/transforms/core.py",
line 1533, in infer_batch_converters
self.fn.output_batch_converter = BatchConverter.from_typehints(
File
"/tmp/tmp.PH23v4EA7w/venv/lib/python3.8/site-packages/apache_beam/typehints/batch.py",
line 90, in from_typehints
raise TypeError(
TypeError: Unable to find BatchConverter for element_type Any and batch_type
<class 'numpy.ndarray'>
```
It's not immediately obvious that this is complaining about the _output_
type of this DoFn. We should:
- Make this error identify the DoFn and if the issue is with input or output
types
- If the issue is with output types and element type is `Any`, we may nudge
them to define `infer_output_type`, as this is a common mistake.
### Issue Priority
Priority: 2
### Issue Component
Component: sdk-py-core
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
